Online evaluation method and system for frequency modulation and peak regulation parameters of power plant side

Abstract

The invention belongs to the technical field of power grid evaluation, and discloses an online evaluation method and system for frequency modulation and peak regulation parameters of a power plant side, and the method comprises the following steps: constructing graph structure data through employing historical operation data; according to the graph structure data, utilizing a graph network model to analyze correlation characteristics among the devices in the power plant, and according to the correlation characteristics among the devices, extracting frequency modulation and peak regulation characteristics through a graph pooling method; dynamically analyzing the extracted frequency modulation and peak regulation features, extracting time sequence features of the frequency modulation and peak regulation features, and constructing a frequency modulation and peak regulation parameter evaluation model in combination with a softmax classifier; and evaluating the real-time operation data of each device in the power plant by using the frequency modulation and peak regulation parameter evaluation model, outputting an evaluation result of frequency modulation and peak regulation parameters, and performing corresponding operation strategy adjustment according to the evaluation result. According to the invention, accurate, timely and effective support can be provided for operation strategy adjustment of the power plant.

Description

Technical Field

[0001] The present invention relates to the technical field of power grid evaluation, and in particular to an online evaluation method and system for frequency and peak regulation parameters at a power plant side. Background Art

[0002] The safe and stable operation of the power system plays a vital role in industrial production and economic development. With the continuous expansion of the power system and the gradual formation of smart grids, the mutual connection and interaction between units and power grids are becoming increasingly close. The requirements of the power grid for the auxiliary services provided by grid-connected units have become more stringent, especially in key indicators such as primary frequency regulation and automatic generation control (AGC), and the assessment and evaluation standards have become clearer, more detailed and more stringent.

[0003] In recent years, the domestic auxiliary service market has gradually transformed, and the performance assessment standards for grid-connected generators in various regions have become increasingly stringent. However, at present, dispatching agencies usually monitor and assess the frequency and peak-shaving performance of power plants through remote control systems. The traditional frequency and peak-shaving assessment method is mainly based on offline data analysis. Although it has certain value in historical data summary and trend analysis, it cannot meet the needs of real-time operation assessment. The lack of dynamic analysis capabilities for real-time data makes the assessment results lag behind and cannot reflect the actual operating status of the unit in a timely manner, resulting in the timeliness and effectiveness of frequency and peak-shaving strategy adjustments being restricted.

[0004] Therefore, how to provide an online evaluation method and system for frequency and peak regulation parameters at the power plant side is a problem that needs to be solved urgently. Summary of the invention

[0005] The embodiment of the present invention provides an online evaluation method and system for frequency and peak regulation parameters at a power plant side to solve the problems in the prior art.

[0006] In order to have a basic understanding of some aspects of the disclosed embodiments, a brief summary is given below. This summary is not a general review, nor is it intended to identify key / important components or to describe the scope of protection of these embodiments. Its only purpose is to present some concepts in a simple form as a prelude to the detailed description that follows.

[0007] According to a first aspect of an embodiment of the present invention, there is provided an online evaluation method for frequency and peak regulation parameters at a power plant side.

[0008] In one embodiment, an online evaluation method for frequency and peak regulation parameters at a power plant side includes the following steps:

[0009] Collect historical operation data of each device in the power plant, and use the historical operation data to construct graph structure data based on the physical and logical connection relationships of the devices;

[0010] According to the graph structure data, the graph network model is used to analyze the correlation characteristics between the various devices in the power plant, and the frequency and peak regulation characteristics are extracted through the graph pooling method based on the correlation characteristics between the various devices;

[0011] Based on the long short-term memory network model, the extracted FM peak modulation features are dynamically analyzed, the temporal features of FM peak modulation features are extracted, and the FM peak modulation parameter evaluation model is constructed in combination with the softmax classifier;

[0012] The frequency regulation and peak regulation parameter evaluation model is used to evaluate the real-time operating data of each equipment in the power plant, output the evaluation results of the frequency regulation and peak regulation parameters, and make corresponding operating strategy adjustments based on the evaluation results.

[0013] In one embodiment, the collecting of historical operation data of each device in the power plant and constructing graph structure data using the historical operation data based on the physical and logical connection relationships of the devices includes the following steps:

[0014] Collecting the operating data of each device in the power plant collected in advance by sensors and performing preprocessing, wherein the preprocessing includes removing abnormal values, filling missing values ​​and data normalization;

[0015] Analyze the physical and logical connection relationships of devices, and use nodes and edges in graph theory to represent devices and connections between devices to obtain a device connection graph;

[0016] The preprocessed operation data is embedded into the connection graph to form graph structure data including the equipment operation status and connection relationship.

[0017] In one embodiment, the method of analyzing the correlation characteristics between the devices in the power plant using a graph network model according to the graph structure data, and extracting the frequency modulation and peak modulation features by a graph pooling method according to the correlation characteristics between the devices includes the following steps:

[0018] Based on the pre-set graph network model, graph convolution operation is performed on the graph structure data, the neighbor characteristics of each node are calculated through the adjacency matrix and the node feature matrix, and a preliminary node feature matrix is ​​generated;

[0019] Through multi-layer graph convolution layers, the correlation characteristics between devices are extracted and a node feature matrix containing the correlation characteristics is generated;

[0020] The preliminary node feature matrix and the node feature matrix containing associated characteristics are fused, and the graph pooling method is used to reduce the dimension and extract features of the fused node feature matrix. Based on the extracted features, the device association characteristics related to frequency modulation and peak regulation are identified, and the frequency modulation and peak regulation features are extracted.

[0021] In one embodiment, the graph structure data is subjected to graph convolution operation based on a preset graph network model, neighbor characteristics of each node are calculated through an adjacency matrix and a node feature matrix, and a preliminary node feature matrix is ​​generated, including the following steps:

[0022] Construct an adjacency matrix and a node feature matrix according to the graph structure data, and normalize the adjacency matrix to obtain a normalized adjacency matrix;

[0023] According to the normalized adjacency matrix and node feature matrix, the neighbor characteristics of each node are aggregated through the graph convolution calculation formula, and the node feature matrix is ​​updated;

[0024] A nonlinear activation function is applied to the updated node feature matrix to generate a preliminary node feature matrix.

[0025] In one embodiment, the graph convolution calculation formula is:

[0026] H (l=1) =σ(AH (l) W (l) );

[0027] In the formula, H (l=1) represents the node feature matrix of the l+1th layer, H (l) represents the node feature matrix of the lth layer, σ represents the activation function, A represents the normalized adjacency matrix, W (l) Represents the trainable weight matrix of layer l.

[0028] In one embodiment, the preliminary node feature matrix and the node feature matrix containing the associated characteristics are fused, and the fused node feature matrix is ​​reduced in dimension and extracted in features using a graph pooling method, and the device associated characteristics related to frequency modulation and peak regulation are identified based on the extracted features. Extracting the frequency modulation and peak regulation features includes the following steps:

[0029] The preliminary node feature matrix and the node feature matrix containing associated characteristics are concatenated according to the feature dimension to generate a fused node feature matrix;

[0030] Use graph pooling method to reduce dimension and extract features from the fused node feature matrix;

[0031] Based on the extracted features and combined with the domain knowledge of power plant operation, the equipment-related characteristics related to frequency and peak regulation are identified, and the frequency and peak regulation features are extracted.

[0032] In one embodiment, the graph pooling method adopts a flat graph pooling method, a layered graph pooling method, a global pooling method or a spatial pyramid pooling method.

[0033] In one embodiment, the method of dynamically analyzing the extracted frequency modulation peak modulation features based on the long short-term memory network model, extracting the time series features of the frequency modulation peak modulation features, and building a frequency modulation peak modulation parameter evaluation model in combination with a softmax classifier includes the following steps:

[0034] The extracted frequency modulation and peak modulation features are sorted in time order to obtain time series features;

[0035] Input the time series features into the long short-term memory network model, and use the long short-term memory network model to extract the global temporal characteristics of the time series features;

[0036] The global temporal characteristics of the time series features are input into the fully connected layer, and the feature transformation is performed. The feature transformation results are input into the softmax classifier, and the probability distribution of each category is input;

[0037] Based on the probability distribution of each category, a frequency modulation and peak modulation parameter evaluation model is constructed and trained using a pre-labeled training dataset.

[0038] In one embodiment, the step of inputting the time series features into the long short-term memory network model and extracting the global temporal characteristics of the time series features using the long short-term memory network model comprises the following steps:

[0039] Construct a long short-term memory network model and use the time series features as input to the long short-term memory network model;

[0040] Extract the hidden state sequence of time series features through the long short-term memory network model;

[0041] The attention mechanism is used to perform weighted summation on the hidden state sequence of time steps to obtain the global temporal characteristics of the time series features.

[0042] According to a second aspect of an embodiment of the present invention, there is provided an online evaluation system for frequency and peak regulation parameters at a power plant side.

[0043] In one embodiment, the online evaluation system for frequency and peak regulation parameters at the power plant side includes:

[0044] The graph structure data construction module is used to collect the historical operation data of each device in the power plant and construct graph structure data using the historical operation data based on the physical and logical connection relationship of the devices;

[0045] The frequency modulation and peak modulation feature extraction module is used to analyze the correlation characteristics between various devices in the power plant based on the graph structure data using the graph network model, and extract the frequency modulation and peak modulation features through the graph pooling method based on the correlation characteristics between various devices;

[0046] An evaluation model building module is used to dynamically analyze the extracted frequency modulation and peak modulation features based on the long short-term memory network model, extract the time series features of the frequency modulation and peak modulation features, and build a frequency modulation and peak modulation parameter evaluation model in combination with a softmax classifier;

[0047] The frequency regulation and peak regulation parameter evaluation module is used to evaluate the real-time operation data of each equipment in the power plant using the frequency regulation and peak regulation parameter evaluation model, output the evaluation results of the frequency regulation and peak regulation parameters, and make corresponding operation strategy adjustments based on the evaluation results.

[0048] According to a third aspect of an embodiment of the present invention, a computer device is provided.

[0049] In some embodiments, the computer device includes a memory and a processor, the memory stores a computer program, and the processor implements the steps of the above method when executing the computer program.

[0050] According to a fourth aspect of embodiments of the present invention, a computer-readable storage medium is provided.

[0051] In one embodiment, the computer-readable storage medium stores a computer program, and the computer program implements the steps of the above method when executed by a processor.

[0052] The technical solution provided by the embodiment of the present invention may have the following beneficial effects:

[0053] The present invention collects historical operating data of various equipment in the power plant and constructs graph structure data based on the physical and logical connection relationships of the equipment. It can comprehensively and systematically reflect the correlation and operating status of various equipment in the power plant. By analyzing the correlation characteristics between equipment through a graph network model, it can deeply explore the mutual influence and dependency relationship between equipment, and provide richer information for the evaluation of frequency and peak regulation parameters. The graph pooling method is used to extract frequency and peak regulation features, which can effectively reduce the dimension of the data while retaining key information and improving the efficiency and accuracy of feature extraction. In the process of feature extraction and evaluation, the method combines the domain knowledge of power plant operation, can more accurately identify the equipment correlation characteristics related to frequency and peak regulation, improve the pertinence and practicality of the evaluation, and can provide accurate, timely and effective support for the adjustment of the power plant's operating strategy.

[0054] It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and, together with the description, serve to explain the principles of the invention.

[0056] Figure 1 is a flow chart showing an online evaluation method for frequency and peak regulation parameters at a power plant side according to an exemplary embodiment;

[0057] Figure 2 It is a principle block diagram of an online evaluation system for frequency and peak regulation parameters at a power plant side according to an exemplary embodiment;

[0058] Figure 3 The figure is a schematic diagram showing the structure of a computer device according to an exemplary embodiment. DETAILED DESCRIPTION

[0059] The following description and accompanying drawings fully illustrate the specific embodiments of this article so that those skilled in the art can practice them. Parts and features of some embodiments may be included in or replace parts and features of other embodiments. The scope of the embodiments of this article includes the entire scope of the claims, as well as all available equivalents of the claims. Herein, the terms "first", "second", etc. are only used to distinguish one element from another, without requiring or implying any actual relationship or order between these elements. In fact, the first element can also be called the second element, and vice versa. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that the structure, device or equipment including a series of elements includes not only those elements, but also other elements that are not explicitly listed, or also include elements inherent to such structure, device or equipment. In the absence of more restrictions, the elements defined by the sentence "including one..." do not exclude the existence of other identical elements in the structure, device or equipment including the elements. Each embodiment is described in a progressive manner herein, and each embodiment focuses on the differences from other embodiments, and the same and similar parts between the embodiments can be referred to each other.

[0060] The terms "longitudinal", "lateral", "upper", "lower", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside", etc. in this document indicate the orientation or position relationship based on the orientation or position relationship shown in the drawings, and are only for the convenience of describing this document and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operate in a specific orientation, and therefore cannot be understood as a limitation on the present invention. In the description of this document, unless otherwise specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a mechanical connection or an electrical connection, it can also be the internal communication of two elements, it can be a direct connection, or it can be an indirect connection through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms can be understood according to specific circumstances.

[0061] As used herein, the term "plurality" means two or more than two, unless otherwise specified.

[0062] In this document, the character " / " indicates that the preceding and following objects are in an "or" relationship. For example, A / B means: A or B.

[0063] In this article, the term "and / or" is a description of the association relationship between objects, indicating that three relationships may exist. For example, A and / or B means: A or B, or, A and B.

[0064] It should be understood that, although the various steps in the flow chart are displayed in sequence according to the indication of the arrows, these steps are not necessarily executed in sequence in the order indicated by the arrows. Unless there is a clear description in this article, the execution of these steps is not strictly limited in order, and these steps can be executed in other orders. Moreover, at least a portion of the steps in the figure may include multiple sub-steps or multiple stages, and these sub-steps or stages are not necessarily executed at the same time, but can be executed at different times, and the execution order of these sub-steps or stages is not necessarily to be carried out in sequence, but can be executed in turn or alternately with other steps or at least a portion of the sub-steps or stages of other steps.

[0065] Each module in the device or system of the present application can be implemented in whole or in part by software, hardware, or a combination thereof. The above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to the above modules.

[0066] In the absence of conflict, the embodiments of the present invention and the features of the embodiments may be combined with each other.

[0067] Figure 1 An embodiment of the online evaluation method for frequency and peak regulation parameters at the power plant side of the present invention is shown.

[0068] In this optional embodiment, the online evaluation method for frequency and peak regulation parameters at the power plant side includes the following steps:

[0069] Step S101, collecting historical operation data of each device in the power plant, and constructing graph structure data using the historical operation data based on the physical and logical connection relationships of the devices;

[0070] In this optional embodiment, the collecting of historical operation data of each device in the power plant and constructing graph structure data using the historical operation data based on the physical and logical connection relationships of the devices includes the following steps:

[0071] Collecting the operating data of each device in the power plant collected in advance by sensors and performing preprocessing, wherein the preprocessing includes removing abnormal values, filling missing values ​​and data normalization;

[0072] Analyze the physical and logical connection relationships of devices, and use nodes and edges in graph theory to represent devices and connections between devices to obtain a device connection graph;

[0073] Specifically, the physical connection relationship can be based on the power plant equipment layout and engineering design drawings to identify the physical connection paths between devices, such as the direct connection between the transformer and the switchgear, the energy transfer path between the power supply equipment and the load equipment, etc. The logical connection relationship can be analyzed by controlling and signaling relationships between devices, such as the control dependency relationship between the host and auxiliary equipment, and the logical dependency relationship between upstream and downstream equipment.

[0074] The preprocessed operation data is embedded into the connection graph to form graph structure data including the equipment operation status and connection relationship.

[0075] Step S102, according to the graph structure data, using the graph network model to analyze the correlation characteristics between the various devices in the power plant, and according to the correlation characteristics between the various devices, extracting frequency and peak regulation features through a graph pooling method;

[0076] In this optional embodiment, the method of analyzing the correlation characteristics between the devices in the power plant using the graph network model according to the graph structure data, and extracting the frequency modulation and peak modulation features by the graph pooling method according to the correlation characteristics between the devices includes the following steps:

[0077] Based on the pre-set graph network model, graph convolution operation is performed on the graph structure data, the neighbor characteristics of each node are calculated through the adjacency matrix and the node feature matrix, and a preliminary node feature matrix is ​​generated;

[0078] Through multi-layer graph convolution layers, the correlation characteristics between devices are extracted and a node feature matrix containing the correlation characteristics is generated;

[0079] The preliminary node feature matrix and the node feature matrix containing associated characteristics are fused, and the graph pooling method is used to reduce the dimension and extract features of the fused node feature matrix. Based on the extracted features, the device association characteristics related to frequency modulation and peak regulation are identified, and the frequency modulation and peak regulation features are extracted.

[0080] In this optional embodiment, the graph structure data is subjected to graph convolution operation based on a preset graph network model, neighbor characteristics of each node are calculated through an adjacency matrix and a node feature matrix, and a preliminary node feature matrix is ​​generated, including the following steps:

[0081] Construct an adjacency matrix and a node feature matrix according to the graph structure data, and normalize the adjacency matrix to obtain a normalized adjacency matrix;

[0082] Specifically, the adjacency matrix is ​​a representation of graph structure data, which is used to describe the connection relationship between nodes in the graph. In the adjacency matrix, the rows and columns represent the nodes in the graph, and the elements in the matrix represent the connection strength or whether there is a connection between the nodes; for undirected graphs, the adjacency matrix is ​​symmetric; for directed graphs, the adjacency matrix is ​​not necessarily symmetric.

[0083] The node feature matrix is ​​used to describe the characteristics or attributes of each node in the graph; each row of the matrix represents a node, and each column represents a feature dimension; the node feature matrix can contain various information of the node, such as physical state, operating state, configuration parameters, etc.

[0084] According to the normalized adjacency matrix and node feature matrix, the neighbor characteristics of each node are aggregated through the graph convolution calculation formula, and the node feature matrix is ​​updated;

[0085] The graph convolution calculation formula is:

[0086] H (l=1) =σ(AH (l) W (l) );

[0087] In the formula, H (l=1) represents the node feature matrix of the l+1th layer, H (l) represents the node feature matrix of the lth layer, σ represents the activation function, A represents the normalized adjacency matrix, W (l) Represents the trainable weight matrix of layer l.

[0088] A nonlinear activation function is applied to the updated node feature matrix to generate a preliminary node feature matrix.

[0089] It should be noted that the nonlinear activation function is a function used to introduce nonlinear characteristics in the neural network; common nonlinear activation functions include ReLU (Rectified Linear Unit), Sigmoid, Tanh, etc.; the role of the nonlinear activation function is to perform a nonlinear transformation on the input features to generate a new feature representation.

[0090] Specifically, the updated node feature matrix is ​​used as input and transformed by a nonlinear activation function. The activation function performs nonlinear processing on each element in the matrix to generate new element values. The matrix processed by the nonlinear activation function is the preliminary node feature matrix.

[0091] In this optional embodiment, the preliminary node feature matrix and the node feature matrix containing the associated characteristics are fused, and the fused node feature matrix is ​​reduced in dimension and extracted in features using a graph pooling method, and the device associated characteristics related to frequency modulation and peak regulation are identified based on the extracted features. Extracting the frequency modulation and peak regulation features includes the following steps:

[0092] The preliminary node feature matrix and the node feature matrix containing associated characteristics are concatenated according to the feature dimension to generate a fused node feature matrix;

[0093] Specifically, when concatenating two matrices according to the feature dimension, assuming that the shape of the preliminary node feature matrix is ​​[N, F1], and the shape of the node feature matrix containing the associated characteristics is [N, F2], then the shape of the concatenated fusion node feature matrix is ​​[N, F1+F2].

[0094] Use graph pooling method to reduce dimension and extract features from the fused node feature matrix;

[0095] Based on the extracted features and combined with the domain knowledge of power plant operation, the equipment-related characteristics related to frequency and peak regulation are identified, and the frequency and peak regulation features are extracted.

[0096] Specifically, the domain knowledge of power plant operation includes equipment types and functions, operating parameters and status, and operating modes and strategies.

[0097] In this optional embodiment, the graph pooling method adopts a flat graph pooling method, a layered graph pooling method, a global pooling method or a spatial pyramid pooling method.

[0098] Step S103, dynamically analyzing the extracted frequency modulation and peak modulation features based on the long short-term memory network model, extracting the time series features of the frequency modulation and peak modulation features, and building a frequency modulation and peak modulation parameter evaluation model in combination with a softmax classifier;

[0099] In this optional embodiment, the method of dynamically analyzing the extracted frequency modulation peak modulation features based on the long short-term memory network model, extracting the time series features of the frequency modulation peak modulation features, and constructing a frequency modulation peak modulation parameter evaluation model in combination with a softmax classifier includes the following steps:

[0100] The extracted frequency modulation and peak modulation features are sorted in time order to obtain time series features;

[0101] Specifically, assuming there are B time steps and each time step has M features, the time series features can be represented as a B×M matrix.

[0102] Input the time series features into the long short-term memory network model, and use the long short-term memory network model to extract the global temporal characteristics of the time series features;

[0103] The global temporal characteristics of the time series features are input into the fully connected layer, and the feature transformation is performed. The feature transformation results are input into the softmax classifier, and the probability distribution of each category is input;

[0104] Based on the probability distribution of each category, a frequency modulation and peak modulation parameter evaluation model is constructed and trained using a pre-labeled training dataset.

[0105] Specifically, the LSTM model (Long Short-Term Memory Network model), the fully connected layer, and the softmax classifier are jointly trained using the training data set. The model parameters are updated through the back-propagation algorithm and optimizer (such as Adam, SGD, etc.), making the model's prediction results more accurate.

[0106] In this optional embodiment, the step of inputting the time series features into the long short-term memory network model and extracting the global temporal characteristics of the time series features using the long short-term memory network model comprises the following steps:

[0107] Construct a long short-term memory network model and use the time series features as input to the long short-term memory network model;

[0108] Specifically, building a long short-term memory network model includes:

[0109] Define the model structure: Determine the input dimension, hidden layer dimension (that is, the number of LSTM units), and output dimension of the LSTM model. The input dimension should match the dimension of the time series features.

[0110] Initialize model parameters: including parameters such as weights and biases, which will be optimized during the training process.

[0111] Select activation function: sigmoid and tanh are commonly used as activation functions in LSTM to process input and hidden states.

[0112] Extract the hidden state sequence of time series features through the long short-term memory network model;

[0113] It should be noted that each unit in the LSTM model will gradually process each time step in the input sequence, and update the hidden state of the current time step according to the current input and the hidden state of the previous time step. The LSTM model will output a hidden state sequence, in which each hidden state contains all the historical information up to the current time step.

[0114] The attention mechanism is used to perform weighted summation on the hidden state sequence of time steps to obtain the global temporal characteristics of the time series features.

[0115] Step S104, using the frequency regulation and peak regulation parameter evaluation model to evaluate the real-time operating data of each device in the power plant, outputting the evaluation results of the frequency regulation and peak regulation parameters, and adjusting the corresponding operating strategy according to the evaluation results.

[0116] Figure 2 An embodiment of the online evaluation system for frequency and peak regulation parameters at the power plant side of the present invention is shown.

[0117] In this optional embodiment, the online evaluation system for frequency and peak regulation parameters at the power plant side includes:

[0118] The graph structure data construction module 201 is used to collect the historical operation data of each device in the power plant, and construct the graph structure data using the historical operation data based on the physical and logical connection relationship of the devices;

[0119] The frequency modulation and peak modulation feature extraction module 202 is used to analyze the correlation characteristics between various devices in the power plant according to the graph structure data using the graph network model, and extract the frequency modulation and peak modulation features through the graph pooling method according to the correlation characteristics between various devices;

[0120] An evaluation model building module 203 is used to dynamically analyze the extracted frequency modulation and peak modulation features based on a long short-term memory network model, extract the time series features of the frequency modulation and peak modulation features, and build a frequency modulation and peak modulation parameter evaluation model in combination with a softmax classifier;

[0121] The frequency regulation and peak regulation parameter evaluation module 204 is used to evaluate the real-time operation data of each device in the power plant using the frequency regulation and peak regulation parameter evaluation model, output the evaluation results of the frequency regulation and peak regulation parameters, and make corresponding operation strategy adjustments according to the evaluation results.

[0122] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, a memory and a network interface connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer device is used to store static information and dynamic information data. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, the steps in the above method embodiment are implemented.

[0123] Those skilled in the art will understand that Figure 3The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present invention, and does not constitute a limitation on the computer device to which the solution of the present invention is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0124] In addition, the present invention also provides a computer device, including a memory and a processor, wherein a computer program is stored in the memory, and the processor implements the steps in the above method embodiment when executing the computer program.

[0125] In addition, the present invention further provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, the steps in the above method embodiment are implemented.

[0126] Those skilled in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be completed by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided by the present invention can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory or optical memory, etc. Volatile memory can include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM can be in various forms, such as static random access memory (SRAM) or dynamic random access memory (DRAM).

[0127] The present invention is not limited to the structures which have been described above and shown in the drawings, and various modifications and changes may be made without departing from the scope thereof. The scope of the present invention is limited only by the appended claims.

Claims

1. An online evaluation method for frequency and peak regulation parameters at a power plant side, characterized in that: The method comprises the following steps: Collect historical operation data of each device in the power plant, and use the historical operation data to construct graph structure data based on the physical and logical connection relationships of the devices; According to the graph structure data, the graph network model is used to analyze the correlation characteristics between the various devices in the power plant, and the frequency and peak regulation characteristics are extracted through the graph pooling method based on the correlation characteristics between the various devices; Based on the long short-term memory network model, the extracted FM peak modulation features are dynamically analyzed, the temporal features of FM peak modulation features are extracted, and the FM peak modulation parameter evaluation model is constructed in combination with the softmax classifier; The frequency regulation and peak regulation parameter evaluation model is used to evaluate the real-time operating data of each equipment in the power plant, output the evaluation results of the frequency regulation and peak regulation parameters, and make corresponding operating strategy adjustments based on the evaluation results.

2. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 1 is characterized in that: The collecting of historical operation data of each device in the power plant and constructing graph structure data using the historical operation data based on the physical and logical connection relationship of the devices includes the following steps: Collecting the operating data of each device in the power plant collected in advance by sensors and performing preprocessing, wherein the preprocessing includes removing abnormal values, filling missing values ​​and normalizing data; Analyze the physical and logical connection relationships of devices, and use nodes and edges in graph theory to represent devices and connections between devices to obtain a device connection graph; The preprocessed operation data is embedded into the connection graph to form graph structure data including the equipment operation status and connection relationship.

3. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 2 is characterized in that: The method of analyzing the correlation characteristics between the devices in the power plant by using the graph network model according to the graph structure data and extracting the frequency and peak regulation features by using the graph pooling method according to the correlation characteristics between the devices includes the following steps: Based on the pre-set graph network model, graph convolution operation is performed on the graph structure data, the neighbor characteristics of each node are calculated through the adjacency matrix and the node feature matrix, and a preliminary node feature matrix is ​​generated; Through multi-layer graph convolution layers, the correlation characteristics between devices are extracted and a node feature matrix containing the correlation characteristics is generated; The preliminary node feature matrix and the node feature matrix containing associated characteristics are fused, and the graph pooling method is used to reduce the dimension and extract features of the fused node feature matrix. Based on the extracted features, the device association characteristics related to frequency modulation and peak regulation are identified, and the frequency modulation and peak regulation features are extracted.

4. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 3 is characterized in that: The method of performing graph convolution operation on graph structure data based on a preset graph network model, calculating neighbor characteristics of each node through an adjacency matrix and a node feature matrix, and generating a preliminary node feature matrix includes the following steps: Construct an adjacency matrix and a node feature matrix according to the graph structure data, and normalize the adjacency matrix to obtain a normalized adjacency matrix; According to the normalized adjacency matrix and node feature matrix, the neighbor characteristics of each node are aggregated through the graph convolution calculation formula, and the node feature matrix is ​​updated; A nonlinear activation function is applied to the updated node feature matrix to generate a preliminary node feature matrix.

5. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 4 is characterized in that: The graph convolution calculation formula is: H (l=1) =σ(AH (l) W (l) ); In the formula, H (l=1) represents the node feature matrix of the l+1th layer, H (l) represents the node feature matrix of the lth layer, σ represents the activation function, A represents the normalized adjacency matrix, W (l) Represents the trainable weight matrix of layer l.

6. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 3 is characterized in that: The preliminary node feature matrix and the node feature matrix containing the associated characteristics are fused, and the fused node feature matrix is ​​reduced in dimension and extracted in features by using the graph pooling method, and the device associated characteristics related to frequency modulation and peak regulation are identified according to the extracted features. Extracting the frequency modulation and peak regulation features includes the following steps: The preliminary node feature matrix and the node feature matrix containing associated characteristics are concatenated according to the feature dimension to generate a fused node feature matrix; Use graph pooling method to reduce dimension and extract features from the fused node feature matrix; Based on the extracted features and combined with the domain knowledge of power plant operation, the equipment-related characteristics related to frequency and peak regulation are identified, and the frequency and peak regulation features are extracted.

7. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 6 is characterized in that: The graph pooling method adopts a flat graph pooling method, a layered graph pooling method, a global pooling method or a spatial pyramid pooling method.

8. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 1 is characterized in that: The method of dynamically analyzing the extracted frequency modulation peak modulation features based on the long short-term memory network model, extracting the time series features of the frequency modulation peak modulation features, and building a frequency modulation peak modulation parameter evaluation model in combination with a softmax classifier includes the following steps: The extracted frequency modulation and peak modulation features are sorted in time order to obtain time series features; Input the time series features into the long short-term memory network model, and use the long short-term memory network model to extract the global temporal characteristics of the time series features; The global temporal characteristics of the time series features are input into the fully connected layer, and the feature transformation is performed. The feature transformation results are input into the softmax classifier, and the probability distribution of each category is input; Based on the probability distribution of each category, a frequency modulation and peak modulation parameter evaluation model is constructed and trained using a pre-labeled training dataset.

9. The online evaluation method for frequency and peak regulation parameters at the power plant side according to claim 8 is characterized in that: The step of inputting the time series features into the long short-term memory network model and extracting the global time series characteristics of the time series features using the long short-term memory network model comprises the following steps: Construct a long short-term memory network model and use the time series features as input to the long short-term memory network model; Extract the hidden state sequence of time series features through the long short-term memory network model; The attention mechanism is used to perform weighted summation on the hidden state sequence of time steps to obtain the global temporal characteristics of the time series features.

10. An online evaluation system for frequency and peak regulation parameters at the power plant side, characterized in that: The system includes: The graph structure data construction module is used to collect the historical operation data of each device in the power plant and construct graph structure data using the historical operation data based on the physical and logical connection relationship of the devices; The frequency modulation and peak modulation feature extraction module is used to analyze the correlation characteristics between various devices in the power plant based on the graph structure data using the graph network model, and extract the frequency modulation and peak modulation features through the graph pooling method based on the correlation characteristics between various devices; An evaluation model building module is used to dynamically analyze the extracted frequency modulation and peak modulation features based on the long short-term memory network model, extract the time series features of the frequency modulation and peak modulation features, and build a frequency modulation and peak modulation parameter evaluation model in combination with a softmax classifier; The frequency regulation and peak regulation parameter evaluation module is used to evaluate the real-time operation data of each equipment in the power plant using the frequency regulation and peak regulation parameter evaluation model, output the evaluation results of the frequency regulation and peak regulation parameters, and make corresponding operation strategy adjustments based on the evaluation results.

Images